Analysis of Genomic Imprinting of UBE3A in Neurons

Angelman syndrome (AS), chromosome 15q11-q13 duplication syndrome (Dup15q), and Prader-Willi syndrome (PWS) are neurodevelopmental disorders associated with dysregulated expression of imprinted genes located within the human 15q11-13 imprinted region. Angelman syndrome is caused by loss-of-function or loss-of-expression of the maternally inherited UBE3A allele; Dup15q syndrome is attributed to maternally inherited copy number gains of UBE3A; and, paternally inherited deletions of the SNORD116 cluster cause PWS. The UBE3A gene is imprinted in the brain with maternal-specific expression and biallelically expressed in all other cell types. The imprint is regulated by expression of the UBE3A antisense transcript (UBE3A-AS), which is expressed only in neurons and imprinted with paternal-specific expression. The UBE3A-AS represents the 3` end of a long polycistronic transcript that includes the SNORD116 and SNORD115 gene clusters. Thus, the genes causing AS, Dup15q, and PWS are transcriptionally linked; however, the functional significance of the neuron specific imprint is largely unknown. In this dissertation, it was hypothesized that imprinting of UBE3A evolved as a mechanism to negatively regulate UBE3A protein levels in neurons. This hypothesis was tested by examining allelic expression patterns and associated protein levels of the mouse 7c imprinted region, the orthologous region of human 15q11-q13.

Analyses revealed that imprinted expression of Ube3a in the brain resulted in elevated RNA and protein levels compared to tissues where Ube3a was biallelically expressed. Likewise, Snord116, Snord115, and Ube3a-AS transcripts were highly expressed in the brain. The elevated Ube3a protein levels in the brain were due to increased maternal-allelic expression during neurogenesis concurrent with paternal-allelic suppression. Analysis of UBE3A expression in the opossum, a metatherian mammal lacking an orthologous imprinted region, showed that the UBE3A imprint did not evolve to negatively regulate UBE3A protein levels in the brain. Extensive alternative splicing of Ube3a-AS was detected in the brain, which generated at least two transcripts containing novel open reading frames. Novel Ube3a alternatively spliced transcripts were also identified in the brain. Collectively, these data reject the hypothesis that the UBE3A imprint evolved to negatively regulate UBE3A protein levels in the brain; instead, they suggest that the UBE3A imprint may allow co-expression of the UBE3A and SNORD gene cluster in neurons, which may also facilitate or regulate the expression of novel brain-specific UBE3A transcripts.